Dealing with Complexity LO6578

Peter Marks (marks@halcyon.com)
Wed, 10 Apr 1996 17:08:08 -0700 (PDT)

Replying to LO6568 --

Richard,
> The new distinction is different. It's between the "physical world" and
> the "human world." The physical world is governed by laws and
> relationships that are relatively fixed. We understand them only partially
> and there is steady progress in understanding them more fully. But, for
> example, the relative reactivity of metals is quite constant; we have
> learned that aluminum will reduce iron ore in the thermite reaction, and
> we are really quite sure that it will keep doing so. The laws and
> realationships of the physical world are lasting, althought imperfectly
> understood.
> ...
> The big difference is that in the human world, the "laws
> and relationships" we observe are created by our culture, language,
> interactions, our consensual domains created to date. They might change
> greatly if we had been thinking differently, if a different accident had

A similar distinction was made by Herbert Simon in "Sciences of the
Artificial". He was contrasting understanding in the natural sciences
(whose laws have the properties you describe) with understanding in such
arenas as engineering, architecture, (judicial) law, where the critical
*design* aspects are human creations and can *in principal* be changed at
will. Of course this latter flexibility has to be taken with a grain of
salt when people (e.g., customers, citizens) have come to depend on things
working as they currently do, and the resulting interests create all sorts
of social pressures. I suspect that Simon's 'artificial' is (at least) a
significant component of your 'human world'.

-- 

P-) ___o - o Peter Marks <marks@halcyon.com> _-\_<, - _\ /\_ 15307 NE 202nd Street (206)489-0501 (*)/ (*) - (*)^(*) Woodinville, WA 98072 ------------------------------ More comfortable AND faster ... that's REAL technology!

Learning-org -- An Internet Dialog on Learning Organizations For info: <rkarash@karash.com> -or- <http://world.std.com/~lo/>